Review & Outlook

Our take on the investing, financial, & economic themes of the day

The Problem with PE Ratios

3 December, 2015 by Ben O'Brien in Commentary

One of the first things new investors tend to learn is the PE ratio. It is a very quick measure of a company’s relative valuation. To get the PE ratio you simply divide the current share price by one year’s earnings per share. The ratio tells you how much you will pay for each dollar of a company’s earnings. This number, often called the earnings “multiple”, is then compared to the PE ratio of other stocks or of an index. For many individual investors the PE has traditionally been an essential factor in identifying good stocks to buy.

When you look closely at PE ratios, however, you begin to find that the PE is a very blunt tool, and maybe not as useful as it first appears. Relying too much on this ratio can lead to all sorts of investment mistakes and misconceptions. Why is it that despite its more than $50 billion in earnings, Apple has a relatively low PE while the barely profitable Amazon has an astronomical one? While I’m not sure I have the answer to this particular investment conundrum, looking at the limitations of the PE ratio as a valuation tool will help to address this puzzling question.

Valuation versus Pricing

One misconception which pops up often in the financial media is the line of thinking that investors can generally be divided into “value” investors who like out-of-favor, low PE stocks and “growth” investors who tend to pick fast-growing, high PE stocks. I saw an example of this recently in an article on Money magazine’s website, ”Is Warren Buffett Still a Bargain Hunter?” The article observes that Buffett’s Berkshire Hathaway has strayed from its value investing discipline because it has bought some companies with high PE ratios.

Unfortunately things are just not so simple, and in fact neither growth nor value investors pay a great deal of attention to raw PE ratios. If the success of Warren Buffett was simply a matter of picking stocks with low PEs, then investing would be easy. Rather, sophisticated investors look at a multitude of factors when valuing a company.

Some experts such the valuation professor Aswath Damodaran even argue that PE ratios are not truly a measure of “valuation”, but are instead simply a form of “pricing”. Valuation, strictly speaking, refers to the process of estimating a company’s intrinsic value, while pricing or relative valuation is merely a comparison to other assets. The problem with pricing or relative valuation is that it doesn’t work at times when the whole market or whatever stocks or sectors you are using for comparison are also mispriced. In the late 1990’s, for instance, a stock trading at 25 times earnings might be considered cheap compared to the market’s average valuation of 32 times earnings. However, the whole market at that point was out of whack, and so the index’s PE ratio was not a very good measuring stick. Measures of intrinsic value, on the other hand, though not without an element of subjectivity, are generally driven by company fundamentals and are no so reliant on market conditions or investor sentiment.

Which Earnings?

Even if you accept that relative valuation using PE ratios is a worthwhile practice, you face the difficult question of which measure of earnings to use for the E in PE. Many investors use “forward earnings”, however, these are based on an estimate of the coming year. Another option which avoids using any hypotheticals is using the last twelve months earnings. This choice can make a big difference. Cal-Maine, the leading U.S. egg producer, for example, currently has a trailing (last twelve month) PE of 9.3, nearly twice the forward PE of 4.8.

Investors also have to choose which accounting treatment to use for earnings. Earnings tend to fluctuate and are prone to manipulation. There are several measures that vary in their treatments of “extraordinary charges” that often include things like acquisition or restructuring charges. Right now with many U.S. companies you have to decide whether to exclude the impact of currencies which is quite large. The choices of how to treat these issues can result in a very different PE ratio.

The cyclical nature of many stocks also casts doubt on whether one year’s earnings number is representative of a company’s real value. For example, during much of 2015 the last twelve months earnings for energy companies contained the earnings for a period of time when oil was trading around $100 per barrel. Now that oil is around $45 these past earnings don’t seem like a realistic measure of the company’s value anymore. One way of getting around this, which was used by Benjamin Graham, is using an average of five or ten years of past earnings to “normalize” the earnings number. This accounts for the full range of earnings across economic cycles. However, five or ten years ago the business might have been very different, and so this method isn’t foolproof either.

Other Drivers of PE Ratio

Even once you are satisfied that you have a good earnings number in the denominator of the ratio, there are lots more factors to consider. Not all earnings are created equal. Investors will pay more for steady, reliable, growing earnings than they will for unpredictable or cyclical earnings. Earnings that have a high degree of visibility, for instance those that come from long-term contracts or essential products can be more valuable and produce a higher PE than for unpredictable earnings such as those from commodity companies or trendy retailers.

The investor and finance professor Sanjay Bakshi wrote in his “Primer on PE Ratios” that the ratio has eight main drivers. These are 1) stability, 2) growth, 3) dividends, 4) return on invested capital 5) leverage, 6) the proportion of non-operating assets in a company’s asset base 7) investor sentiment regarding the industry and the company and 8) interest rates.

Though I won’t get into discussing all eight factors here, you can see that when you have to consider all of these sort of things, the raw PE ratio does not have a lot of meaning. Even considering these factors is not straightforward. For instance, Bakshi points out, the impact of dividends can work both ways. Paying a dividend generally makes a stock more attractive and thus gives it a higher PE, but when the dividend gets too high, the company no longer has enough earnings left over to invest in growth, and so the PE may fall. This is why Mattel with a dividend yield of 6% has a lower PE than rival toy maker Hasbro, which only has a dividend of 3%. Hasbro has the prospect of greater earnings growth.

The problem with PE ratios, I think, is that they are an attempt to quantify valuation in one simple number, but valuation is an inherently subjective exercise with a number of moving parts.

Amazon versus Apple

So where does this leave us with Amazon and Apple? I’m still not sure I’ve solved the puzzle, but I think it’s a good illustration of the limits of PE ratios. At the beginning of this year Apple had vast profits and a low PE and Amazon was still barely profitable with a huge PE. Who would you have bet on? All signs pointed towards Apple. But this year Apple is up 5%, while Amazon skyrocketed more than 100%.

It seems somehow, despite the fact that Amazon has been around since the 90s, investors have decided to buy into Amazon’s strategy of operating like an early-stage start-up company. For such companies Bakshi’s numbers 1-6 are not as important since the company doesn’t have a track record of earnings to measure, and so number 7, the most subjective factor, becomes all important. This is how venture capitalists operate, taking big bets on future prospects that are very uncertain but should they work out would have a huge payoff.

Apple on the other hand is a fully mature company. As the biggest company in the world it is facing a sort of winner’s curse and running up against the law of large numbers. Even if Apple continues to be highly successful, when growing off of a base of over $600 billion in market cap, they cannot possibly continue their earlier 50+% growth rates. Thus Apple is seen as a company with slowing growth, and investors are worried the company has lost the magic of the Steve Jobs years.

I am not arguing that either of these valuations is correct. The bottom line is that valuation is complicated and somewhat subjective. It is not a purely quantitative exercise, but requires all sorts of speculative judgements on the quality of management, the future of industries and so on. So while PE ratios appear to be a useful shorthand in valuation, their effectiveness in making investment decisions is limited and often misunderstood. As with most rules of thumb, they are no substitute for hard analysis.